Incident History for Replicate

The status page migration maintenance has completed and all affected components are operational.

The status page migration is complete, though DNS propagation may still affect access.

Maintenance to migrate status page capabilities to cloudflarestatus.com is in progress, affecting all components.

Scheduled maintenance to migrate status page capabilities to cloudflarestatus.com, affecting all components.

A100 hardware capacity has been restored and all queues are drained.

Monitoring started; pods are starting up again and working through the backlog.

We're investigating a partial outage where most A100 hardware is automatically cordoned due to possible hardware failure.

All affected models have been scaling correctly for over 30 minutes with no residual prediction queues.

Scaling decisions were delayed nearly 1 hour after a controller failed to schedule on a soft-failed node; the node was cordoned and drained, and metrics are flowing again.

The incident has been resolved after monitoring; services are back to normal.

Monitoring started after significant progress; most traffic is flowing normally, and we are monitoring to ensure full recovery.

We have identified the root cause of a significant degradation and are taking steps to mitigate.